a8b88bc01d48015c22420c4a7aae90ca

This model is a fine-tuned version of albert/albert-large-v1 on the nyu-mll/glue dataset. It achieves the following results on the evaluation set:

• Loss: 0.3651
• Data Size: 1.0
• Epoch Runtime: 184.9100
• Accuracy: 0.8744
• F1 Macro: 0.8744
• Rouge1: 0.8746
• Rouge2: 0.0
• Rougel: 0.8743
• Rougelsum: 0.8741

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 5e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• total_train_batch_size: 32
• total_eval_batch_size: 32
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: constant
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Data Size	Epoch Runtime	Accuracy	F1 Macro	Rouge1	Rouge2	Rougel	Rougelsum
No log	0	0	0.6891	0	3.4954	0.5303	0.4846	0.5303	0.0	0.5303	0.5305
No log	1	3273	0.5550	0.0078	6.2844	0.7414	0.7322	0.7414	0.0	0.7410	0.7412
0.0093	2	6546	0.4878	0.0156	6.4388	0.7719	0.7630	0.7715	0.0	0.7715	0.7717
0.4516	3	9819	0.3727	0.0312	9.2452	0.8518	0.8507	0.8518	0.0	0.8517	0.8517
0.4349	4	13092	0.3584	0.0625	14.9013	0.8526	0.8522	0.8529	0.0	0.8528	0.8526
0.3481	5	16365	0.3817	0.125	26.1150	0.8504	0.8497	0.8504	0.0	0.8504	0.8502
0.3393	6	19638	0.2990	0.25	48.0217	0.8809	0.8808	0.8811	0.0	0.8811	0.8807
0.326	7	22911	0.3134	0.5	93.4835	0.8717	0.8712	0.8717	0.0	0.8717	0.8715
0.3125	8.0	26184	0.3026	1.0	185.6783	0.8779	0.8778	0.8781	0.0	0.8781	0.8779
0.2648	9.0	29457	0.2835	1.0	185.8369	0.8866	0.8865	0.8864	0.0	0.8866	0.8864
0.2234	10.0	32730	0.2961	1.0	185.8549	0.8864	0.8863	0.8864	0.0	0.8862	0.8862
0.1967	11.0	36003	0.3656	1.0	185.2062	0.8844	0.8843	0.8844	0.0	0.8842	0.8844
0.1727	12.0	39276	0.3302	1.0	185.2051	0.8858	0.8858	0.8857	0.0	0.8855	0.8858
0.1363	13.0	42549	0.3651	1.0	184.9100	0.8744	0.8744	0.8746	0.0	0.8743	0.8741

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu128
• Datasets 4.0.0
• Tokenizers 0.22.1